Analysis It was not a very auspicious beginning to the merger of two tech giants. Oracle just couldn't seem to get its phone lines and Webcast in order to do the actual announcement of its $5.6bn acquisition of server maker Sun Microsystems.
Or, perhaps this explains it all. I guess Oracle needed some Sun servers a little more quickly than it might have thought, eh?
With the Webcast unavailable and the phone lines at Oracle's conference call system all snarled up, I was spared the often condescending yet always dulcet tones of Larry Ellison, co-founder and chief executive officer of Oracle, and Jonathan Schwartz, president and chief operating officer at Sun, for which I suppose I am grateful. (Until I have to circle back and listen to the replay of the press conference to finish this article.)
I am also grateful that the month-long saga between IBM, which flirted with the idea of buying Sun for more dough and then less than Oracle is shelling out, and Sun is over. It was painful to contemplate the crashing of the East coast IT giant with the upstart West coast former giant (at least in terms of influence and market capitalization a decade ago). Sun and IBM mixed like oil and water. Oracle and Sun, the two original Silicon Valley IT startups (and no, Hewlett-Packard, oscilloscopes don't count as IT) and the two darlings of the dot-com boom, will mix something more like vinegar and oil - you can whip it up into a colloid, at least.
Sun is a much more natural fit inside Oracle, but it is a deal that should have never had to happen in the first place, and if anything, looking back a decade ago, Sun should have acquired Oracle and then did the run on the application and middleware software space as a much larger and stronger organization.
For Sun, the Oracle acquisition happening today does two things. First and foremost, it stops all the talk about the Bigger Indigo deal, particularly the muttering about lawsuits because Sun turned down the offer that IBM made a few weekends ago for around the same amount of money. With IBM reporting its financial results for the first quarter of 2009 today after Wall Street closes, Big Blue was going to have to say something about the deal, and with Sun reporting its fiscal 2009 third quarter results next Tuesday, people were going to have plenty of time to think about whatever IBM might say. Now, IBM won't say jack about the deal and Sun doesn't have to, either.
That said, if IBM was gambling that Sun would somehow become damaged goods after Big Blue made a run at it, and thereby be busted up, that gamble does not seem to be paying off. Sun and Oracle together will (eventually) be stronger than the two apart. And the IT giant may eventually not only rue the day two decades ago when it decided to get out of the application software business to be a kind of platform Switzerland, but that Oracle has now built up the kind of systems business that IBM once had: an integrated stack of systems, systems software, and application software that comes in a number of different flavors to suit different customers' histories and proclivities. Rather than calling is Sunoracle or Oraclesun, perhaps Big Red is in order. Especially once Oracle acquires Red Hat. But let's not get too far ahead of ourselves.
Oracle is buying Sun to get ahold of two key software assets - the Solaris variant of Unix and the Java language and application environment that Oracle has itself staked its application future upon.
Buying Sun also means getting a whole bunch of Sparc and x64 servers, various disk and tape storage products, a slew of open source projects, and an army of hardware and software engineers and researchers. These are not things that Oracle has needed to date, but it looks like Oracle will be positioning itself as a top-to-bottom systems provider, much as IBM has always been, HP sometimes is, and as Cisco most certainly wants to become.
Oracle has the advantage of having the second-largest application software business in the world, and the Sun-Oracle deal may just do something that has been rumored many times in the past: get IBM to acquire German software giant SAP. There could hardly be a more natural fit for Big Blue than SAP, considering that it was founded by some ex-IBMers back when mainframes were not only cool, but basically the main computing environment available. Then IBM and Oracle would have similar sales pitches, leaving Cisco and HP out in the cold without applications.
How very AS/400-like that would be.
Team Reg will be picking this Sun acquisition apart with scalpels, looking at the implications for servers, storage, operating systems, databases, and such. Oracle will not give out precise details about product roadmaps until the deal closes, possibly this summer, but the company did put out a FAQ file for customers and partners (here), a basic presentation (here), and a statement from Charles Philips, Oracle's co-president (here).
On the call with analysts and journalists (the replay was finally posted), Safra Catz, Oracle's other co-president, said that Oracle would be using a mix of cash and debt to acquire Sun, and that the company believed that the Sun unit could deliver $1.5bn in non-GAAP operating income per year, and then grow from there. She added that Oracle believed that Sun, once integrated into Oracle, would be accretive to earnings by 15 cents per share - making the acquisition more profitable than the acquisitions of PeopleSoft, Siebel Systems, and BEA Systems combined.
Clearly, there are some pretty big layoffs coming to Sun again to make such numbers work. And lest we wonder about Oracle's intent, Katz said that Oracle would combine the software units quickly and that Sun's x64 and Sparc systems would be oriented to their joint, existing enterprise customers.
"We intend to ensure that it is a profitable operating unit within Oracle," Catz said, referring to Sun's Systems Group.
Ellison piped up that in Oracle's opinion, Solaris was the best Unix in the market, and reminded everyone that more Oracle databases are deployed on Solaris than any other operating system, but that Linux is number two on the list and that Oracle's commitment to Linux has not been diminished one iota by the acquisition of Sun. Ellison then added that Oracle would now be able to tightly integrate Oracle databases and the Solaris operating system, and offer customers a "complete integrated computer system, from database to disk."
Sun co-founder and chairman, Scott McNealy, said somewhat unenthusiastically, that the merger was the "next big step" in the two-decade partnership between Sun and Oracle, and that it was a "truly momentous day for the industry" and that Sun was "thrilled to be acquired by Oracle."
Schwartz added that the deal brings to market a "new leader" and an "industry phase change" that collapses many different markets into one player - a sales pitch Sun itself has been using, especially since it acquired MySQL in January 2008 for $1bn. Schwartz added that Sun would be the largest
contributor to supplier of open source software, and then waxed a little poetic. (Inhale.)
"There is no question in my mind that this transaction redefines the industry, redefining the boundaries that have frustrated the industry's ability to solve the problems that customers seek to solve, eliminating the cost and complexity of an industry that focuses on components instead of systems," Schwartz said. "The combined Oracle and Sun will be just that: a systems and software powerhouse."
Philips got to speak last on the short call, and said that Oracle spends a lot of money making sure its databases, middleware, and applications all run properly on a wide variety of different piece parts, but that by "engineering a true system" Oracle would be able to eliminate costs and reduce the total cost of ownership of its systems. He said that Oracle might, for instance, deliver ready-to-deploy servers, "a complete industry in a box" on a Sun-Oracle appliance.
Of course, cutting those software engineering and testing costs Philips mentioned would mean eventually not supporting other non-Oracle stuff. But Philips didn't say this was the plan, and even if it is, Oracle has to say it isn't at least for the near term. ®